Coding

Coding skills help developers generate, update, and enhance MuleSoft implementation code using natural language prompts. These capabilities streamline tasks such as API spec generation, DataWeave transformation scripting, and full-flow creation, with support for working from scratch, with repositories, or local project uploads.

Key Capabilities

• Prompt-Driven Code Generation: Turn plain English descriptions into Mule flows, components, and configuration code.
• Flexible Input Modes: Start with no repo or connect an existing repository, or upload a local MuleSoft project folder.
• Multi-tasking: These specialized handle everything from API spec creation to DataWeave scripting.
• Contextual Understanding: Generate output that aligns with your Mule/Java version, runtime config, and business rules.
• Interactive Feedback Loop: Review generated code, provide feedback, and approve the final version before applying.

---

Featured Coding Capabilities

1. API Spec Generator

Generates RAML 1.0 or OAS 3.0 specifications from natural language prompts.

• Supports three input modes: No Repository, With Repository, and Upload from Computer.
• In repo-based modes, API spec generator can analyze existing implementation code to generate or enhance the spec structure.
• Ideal for creating API specifications from scratch or updating specs based on existing MuleSoft projects.

2. Repository Coder

A multi-tasking AI capability designed to help you both build new MuleSoft implementation code from scratch and enhance existing projects.

• Supports all three input modes: No Repository, With Repository, and Upload from Computer.
• Capable of generating complete Mule flows, DataWeave transformations, connector configurations, and testable logic.
• Understands your existing repository and builds intelligently on top of it.
• Ideal for tasks like creating new flows, enhancing existing logic, upgrading connectors or runtimes, and more.

Additional Features (Optional):

• Run MUnit Tests: If required, check the MUnit run option to automatically execute tests after code generation.
• Flow Generation from Diagrams: Optionally upload a sequence or flow diagram to help the repository coder structure your implementation.
• Runtime & VM Configuration: Select Mule/Java runtime versions and provide custom VM arguments, if needed, to match your environment.
• settings.xml Support: Upload a settings.xml file if you need to customize connector resolution or Maven settings.

This capability brings contextual awareness and flexibility, making it an ideal co-pilot for both greenfield development and iterative enhancements.

3. DataWeave Generator

Generates accurate and reusable DataWeave (DWL) scripts based on sample input-output payloads — no prompt required.

• Produces clean, optimized transformations without needing code context or instruction prompts.
• Just select the input format and provide matching input/output samples to generate DWL.
• Great for building new transformations or refining existing ones using real data.

Supported Input Formats:

• XML, JSON, CSV, YAML, TEXT, EDI, EDIFACT

Supported Input Modes:

• No Repository: Generate DWL purely from structured examples.
• With Repository: Reference files from an existing repo to guide generation.
• Upload from Computer: Import a MuleSoft project to add additional context.
• With Mapping Table: Upload a structured CSV mapping table to automatically generate a DataWeave script.

This capability is ideal for building robust data transformations fast, without needing to write DWL manually.

Best Practices

• Be specific in your prompt to get high-quality results (e.g., “Create a RAML spec for a Customer API with GET and POST”).
• Use the Notes field to add naming conventions, rules, or expected behaviors.